[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: unexpected error message after Hausman test for random vs. fixed effects

From   "Alfred Stiglbauer" <[email protected]>
To   <[email protected]>
Subject   st: unexpected error message after Hausman test for random vs. fixed effects
Date   Mon, 13 Oct 2008 14:34:50 +0200

Dear all,

I want to conduct some fairly standard Hausman tests (for various specifications) after -xtreg- testing whether a random effects model is appropriate as opposed to fixed effects.

When the program should perform the first of these tests this results in an error message:

hausman cannot be used with clustered or p-weighted data

I do not understand this. I've neither data with weights nor is a cluster option involved anywhere. (Additionally, the data aren't based on a survey and I have not used any command to instruct Stata to treat them in such a way.)

The regression commands include the "robust" option in both regression models (i. e. xtreg[...], fe and xtreg[...], re). Interestingly, the Hausman test is performed when I do NOT use this option.

Any ideas or suggestions? Thanks for help!


P. S.: The code:

foreach refnorm in "external" "leadership" {;
	foreach realvar in "gdp_fc" "absd_unr_fc" {;
    		foreach model in "fe" "re" {;
    			qui xtreg wagegrowth `refnorm' infl_fc_pm_v2 `realvar_fc' d_1990s d_2000s `reg_sample', robust `model';  		
			estimates store est_`refnorm'_`realvar'_`model';
    	hausman est_`refnorm'_`realvar'_fe est_`refnorm'_`realvar'_re;

Alfred Stiglbauer
Oesterreichische Nationalbank (OeNB)
Abteilung fuer volkswirtschaftliche Analysen - Economic Analysis Division
Otto-Wagner-Platz 3, A-1090  Wien,   A u s t r i a
Phone +43 (0)1 40420, ext. 7435
Mobile +43 (0)664 8580706
Fax +43 (0)1 40420, ext. 7499
[email protected] 

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index